Abstract: The dataset GSE161731 was downloaded from the Gene Expression Omnibus(GEO) database. Variation analysis and Lasso regression were used to search for the key genes which can distinguish early COVID-19 patients. Diagnostic efficacy was evaluated by ROC curve. Biomarkers of different disease groups were confirmed by co-expression network constructed (WGCNA). KEGG enrichment analysis was applied to explain the function of gene modules and related signal pathways. A total of 509 differential genes were screened, and 17 potential diagnostic markers for early-stage COVID-19 patients were identified including MYBL2, PKMYT1, HJURP, TCN2, TTC24, ESPL1, GZMK, RPA3, ATP5F1E, CD2, ZAP70, IL15RA, NFYB, CIAO2A, PLRG1, GPATCH11 and MRPL33. KEGG results showed that these differentially expressed genes were mainly related to cell cycle, T cell immunity, and oxidative phosphorylation. The COVID-19 specific gene modules identified by WGCNA analysis were mainly involved in physiological activities such as protein synthesis and ribosomes oxidative phosphorylation.

Key words: COVID-19, peripheral blood, bioinformatics analysis, transcriptome signature